6. REFLEXIONES FINALES: ¿PERPETUACIÓN O DISMINUCIÓN DE LA
6.1. La Familia Frente a los Factores de Riesgo
6.1.2. Prácticas Educativas que Buscan Responder a los Factores de Riesgo
The P3HT:PCBM devices based on ZnO NW with six different lengths (fabrication process described in 5.2) are investigated for the performance as a function of the NW length. The flat ZnO layer obtained via sol-gel method is used for the reference device, in comparison with ZnO NW devices. It is also used in all the ZnO NW devices as a seed layer for the HT growth.
SEM cross-section characterization of the devices
SEM cross-section images of all the devices are shown in Figure 5.9. It is obvious that the P3HT:PCBM blend is filled completely into the spaces between the nanowires with length
5.4 P3HT:PCBM solar cells based on ZnO NW
from 50 nm to 150 nm. However, the SEM image suggests that the device on 200 nm ZnO NW may have space in the bottom of the array that is not filled. The 250 nm ZnO NW film is not included in this comparison because they have peeled off from the substrate as observed in Figure 5.8f. Both of these two devices are assumed to perform worse than normal, because the charge transport may be interrupted at the interfaces of the blend and the NW by the poor filling, or on the ITO contact due to the peeling off of ZnO NW.
Figure 5.9: SEM cross-section images of blend solar cells based on ZnO seed layer and ZnO NW with
different lengths from 50 nm-250 nm. Blend layer coated on the ZnO NW are much thicker than that on the seed
layer; the ZnO NW devices thickness is almost constant, regardless of the particular ZnO NW length.
It is observed that the blend layers coated on the ZnO NW are much thicker than that on the seed layer, despite their spin coating processes being the same. It indicated that the rough surface of NW may assist to obtain more organic material during the spin coating process. However, the thickness of the entire devices based on ZnO NW is almost constant, about 400 nm for all the solar cells, regardless of the particular ZnO NW length. Therefore, the thickness of the blend layers covered on top of the ZnO NW decreases with the increasing length of NW.
Absorption measurements
Since the exciton generation is induced by the photo injection into the organic conjugated material, the part of solar light absorbed by the ZnO NW does not participate in the excitation. In this case, the absorption of the ZnO NW has to be subtracted from the absorption spectra measured for the whole devices, in order to investigate the influence of light absorption on the
performance of blend-ZnO NW solar cells. 300 400 500 600 700 800 0.0 0.5 1.0 1.5 2.0 ZnO seed ZnNW 50nm ZnNW 80nm ZnNW 100nm ZnNW 150nm ZnNW 200nm ZnNW 250nm
Wavelength(nm)
A
b
s
o
rp
tio
n
(A
.U
.)
Figure 5.10: Absorption of blend layers only coated on ZnO seed layer (black) and ZnO NW with different
lengths: 50 nm (red), 80 nm (green), 100 nm (blue), 150 nm (cyan), 200 nm(magenta), 250 nm(yellow ). The
absorption of blend layer coated on the 50 nm ZnO NW is much higher than that on the seed layer; while it decreases with the length of ZnO NW increasing from 50 nm to 200 nm.
Figure 5.10 shows the absorption results of P3HT:PCBM layers coated on different lengths of ZnO NW. Absorption peaks around 500 nm belong to P3HT, and ones at the 600 nm are induced by PCBM. The peaks shown at 335 nm are attributed to UV light filtered by the ZnO NW with different lengths.
Compared to the blend layer coated on a seed layer only, the one coated on 50 nm ZnO NW shows a significantly improved absorption, which is also the highest absorption among all the blend layers recorded on different ZnO substrates. With the length of ZnO NW increasing from 50 nm to 200 nm, a clear decreasing trend is observed for the absorption of the blend layer coated on those substrates. A possible explanation is given as follows: since the space between the nanowires is relatively small, the blend materials intercalated inside are much
5.4 P3HT:PCBM solar cells based on ZnO NW
less than the part covered on top of ZnO NW. Therefore, the absorption of the blend layer is mainly attributed to the latter part, which has a thickness that decreases strongly depending on the length of the ZnO NW underneath (see Figure 5.9). Thus a reduction in the blend layer absorption follows for longer NW.
However, the device with 250 nm ZnO NW is out of this trend, because the peeling and disordering of its NW (as seen in SEM cross-section images) could seriously influence its absorption and the resulting solar cell performance. Therefore, it should be considered independently from the other ZnO NW devices.
EQE characterizations 300400500600700800 0102030405060 EQE(%) Wavelength(nm) ZnO seed ZnNW 50nm ZnNW 80nm ZnNW 100nm ZnNW 150nm ZnNW 200nm ZnNW 250nm 300 400 500 600 700 800 0 10 20 30 40 50 60
E
Q
E
(%
)
Wavelength(nm)
ZnO seed ZnNW 50nm ZnNW 80nm ZnNW 100nm ZnNW 150nm ZnNW 200nm ZnNW 250nmFigure 5.11: EQE results of blend solar cells based on ZnO seed layer (black) and ZnO NW with different
lengths: 50 nm (red), 80 nm (green), 100 nm (blue), 150 nm (cyan),200 nm (magenta), 250 nm (yellow). The
EQE at 350 nm clearly decreases with the increasing ZnO NW length, while the part at 520 nm decreases with the length of ZnO NW increasing from 50 nm to 200 nm.
To investigate the working mechanisms of the photovoltaic devices under solar illumination, the spectrally resolved current generation is analyzed by EQE measurements. Figure 5.11 shows the EQE characterization results for the blend solar cells on ZnO NW with lengths from 50 nm to 250 nm. The EQE around the wavelength of 350 nm clearly decreases with the increasing ZnO NW length. It is attributed to the filter effect from the absorption of ZnO layers (see section 4.3.5) that can be briefly described as follows: longer ZnO NW can absorb more UV light, but this part of light is filtered out rather than involved in photogeneration. A maximum EQE around 500-600 nm attributed to the absorption of P3HT and PCBM, is observed in all devices. It has a similar dependence on NW length as shown in the absorption results, which decreases with the length of ZnO NW increasing from 50 nm to 200 nm.
Internal quantum efficiency
0102030405060
0.000.050.100.150.200.250.30
EQE/ Absorption
HT growing time (min)
0 10 20 30 40 50 60 0 5 10 15 20 25 30 E Q E / A b s o rp ti o n
HT growing time (min)
Figure 5.12: Internal quantum efficiency of solar cells obtained by dividing EQE by absorption of blend layers that coated on ZnO seed layer (black) and ZnO NW with different lengths: 50 nm (red), 80 nm (green), 100 nm (blue), 150 nm (cyan), 200 nm (magenta), 250 nm (yellow).
To further investigate the dependence of device performance on the ZnO NW length, the influence of differing light absorption caused by the device thickness has to be removed. The
5.4 P3HT:PCBM solar cells based on ZnO NW
internal quantum efficiency (IQE) of devices is obtained by dividing the EQE results by the absorption spectra (Abs) of the blend layers[213]:
IQE(λ) =EQE(λ)/ Abs(λ) (5.1)
As shown in Figure 5.12, the IQE results of devices slightly increase while the NW length increases from 50 nm to 150 nm. It indicates that the exciton collection efficiency might be enhanced by the ZnO NW intercalating the blend system. For the 200 nm and 250 nm NW devices, the opposite trend in the charge transfer could be caused by poor filling of the organic material or nanostructure peeling problems. On the other hand, the significant difference between the devices based on the seed layer and the 50 nm NW indicates that these two layers perform differently in blend-ZnO system for the electron collection. The obviously low IQE of the device based on 50 nm ZnO NW indicates that the high absorption seems to be the main reason for its improved EQE.
As discussed in section 1.2.1, the optical-to-electrical conversion process is the result from:
ŋEQE = ŋAbs · ŋED · ŋCT · ŋCC (5.2)
where ŋAbs is the absorption efficiency, ŋED is the exciton diffusion efficiency, ŋCT is the charge transfer efficiency, and ŋCC is the charge collection efficiency.
Using the equation of 5.1, the IQE is determined by the ŋED, ŋCT and ŋCC. The ŋCT could be negatively influenced by the increasing blend layer thickness, due to the random mixture of donor and accepter [31]; ŋCC, on the other hand, might be improved by the large surface and long pathways for directly electron transport, provided by the NW. The variation of ŋED is not clear: since the exciton diffusion occurs on the interface of P3HT and PCBM, the blend morphology between the NW and seed layer based devices is expected to be different, resulting in different ŋED; on the other hand it might be also effect by the blend layer thickness due to the light intensity loss during the photo injection through this layer [160].
Although the absorption seems to affect most on the result of EQE, the charge transport and collection induced by the NW may influence on the device performance as well. Additionally, the IQE or EQE is not identical to photovoltaic energy conversion; therefore, further investigations on the performance of devices depending on NW length have to be given by I-V characterization.
I-V characterizations
Figure 5.13: Parameters of device performance depending on the length of ZnO NW. (0 nm stands for the
ZnO seed layers) I-V results carried out for blend solar cells based on ZnO seed layer (black) and ZnO NW with
different lengths: 50 nm (red), 80 nm (green), 100 nm (blue), 150 nm (cyan), 200 nm (magenta), 250 nm (yellow).
5.4 P3HT:PCBM solar cells based on ZnO NW
seed layer and ZnO NW with different lengths. Data is acquired under simulated solar illumination and extracted for the six individual pixels of each cell set and the respective standard deviations are calculated. Values are given for solar cells with different thicknesses of ZnO. ZnO substrates Type PCE (%) VOC (V) ISC (mA/cm2) FF (%) RSh (Ω/cm2) RS (Ω/cm2) Seed layer 1.58 0.54 6.48 45.4 348 7.3 ZnO NW 50 nm 1.91 0.57 7.13 47.3 366 6.5 ZnO NW 80 nm 1.62 0.55 6.39 45.5 331 5.8 ZnO NW 100 nm 1.54 0.56 6.31 45.3 343 5.3 ZnO NW 150 nm 1.24 0.52 5.29 45.2 325 7.9 ZnO NW 200 nm 1.07 0.51 4.78 44.1 321 8.5 ZnO NW 250 nm 0.77 0.44 3.55 49.0 642 3.6
Table 5.3: I-V Characteristics results of blend solar cells based on ZnO seed layer and ZnO NW with different lengths from 50 nm-250 nm.
Among all the solar cells with different ZnO layers, the best performance is given by the device with 50 nm ZnO NW, demonstrating values for the VOC, ISC, FF and PCE of 0.57 V, 7.13 mA/ cm2, 0.47, and 1.91 %, respectively. Compared with the solar cell based on a ZnO seed layer, a nearly 20 % improvement in the PCE is shown in 50 nm ZnO NW devices, as a result of increased FF, VOC and ISC. With the length of ZnO NW increasing from 50 nm to 200 nm, both ISC and VOC decrease, but FF, RS and RSh show slightly variations with unclear trends. Furthermore, the 250 nm NW device behaves much differently form the other ones as expected and discussed before.
Corresponding to the discussion about EQE and IQE results, the PCE of devices is mainly attributed to the absorption of blend layers, since it presents a similar NW length dependence
as shown in SEM images and absorption results of blend layers (see Figure 5.9 and 5.10). However, the influences of NW length are not only shown in photocurrent, which is responding to the absorption efficiency, but also present in FF and VOC. It indicates that the NW can influence on the charge transfer of devices as well. This is investigated by comparing the 50 nm ZnO-based device with the flat ZnO solar cell on the I-V curves, giving more information of the charge transport in the devices.
0.0 0.1 0.2 0.3 0.4 0.5 0.6 -8 -7 -6 -5 -4 -3 -2 -1 0 -0.4 -0.2 0.0 0.2 0.4 0.6 0.8 1.0 1E-5 1E-4 1E-3 0.01 0.1 1 10 Bias (V) ZnO seed 50nm ZnO NW
C
u
rr
e
n
t
d
e
n
s
it
y
(
m
A
/c
m
²)
Bias (V)Figure 5.14: I-V curves of blend solar cells based on: a) ZnO seed layer (black square), b) 50 nm ZnO NW
(red circle). Data is acquired under illumination with simulated AM 1.5G solar light (100 mW/cm2).
Figure 5.14 shows the I-V curves of devices based on a ZnO seed layer only and 50 nm ZnO NW. The light I-V curve of the 50 nm ZnO NW solar cell presents a better FF and a steeper slope near the VOC than the seed layer cell, which suggests a lower RS and a higher RSh, as confirmed in Table 5.3. As shown on the dark I-V curves, the device based on 50 nm ZnO NW exihibits a much lower current in back direction compared with the ZnO seed layer device, which may indicate a lower charge carrier recombination rate in the NW cell. All these facts discussed above indicate that the charge transport ability of ZnO NW device is improved, together with the enhanced photocurrent, leading to a better device performance.
5.4 P3HT:PCBM solar cells based on ZnO NW
pathways for the electrons transport when they intercalate into the organic materials; on the other hand, the additional layer of NW grown on flat ZnO films may also provide better hole blocking properties for the device. The reasons for the improved charge transport have to be given by further study, like the light intensity measurements.
Light intensity measurements
Figure 5.15: Light intensity measurement of blend solar cells based on: a) ZnO seed layer (black square), b) 50 nm ZnO NW(red circle).
Figure 5.15 shows the light intensity results of solar cells on a ZnO seed layer only substrate and a 50 nm ZnO NW substrate. It is observed that the VOC of the device on the 50 nm ZnO NW depends linearly on the logarithm of light intensity, which fits the theory of Langevin recombination shown in the equation below [214]:
(5.3)
This formula predicts the right slope S of VOC versus light intensity. The variables in the equation are defined as: Egap is the energy bandgap denoting the energy difference between the
HOMO and LUMO levels, q is the elementary charge, k is Boltzmann‘s constant, T is temperature, γ is the Langevin recombination constant, Ncis the effective density of states P is the dissociation probability of a bound electron-hole pair into free charge carriers, G is the generation rate of bound electron-hole pairs, The generation rate of free charge carriers is then represented by PG, both of which do not depend on intensity [215].
On the other hand, the VOC-light intensity curve for the device on the ZnO seed layer can be fit by two straight lines with different slopes. It indicates that other recombination should be considered in this situation. As discussed in the paper reported by Rappaport et al, [216], the light intensity dependence of the VOC could also be influence by another recombination mechanism, Shockley–Read–Hall (SRH) recombination, which is a trap assisted recombination mechanism [217]. Since the SRH has a different dependence on carrier density than the Langevin recombination [29], it indicates that the significantly increased recombination responding at higher light intensities (>30 mW/cm2) may be the result of these two recombination mechanisms.
The additional trap-assisted recombination is only one reasonable explanation for the low performance of the device based on ZnO seed layer. Other reasons like mobility, structure, and surface properties, as results of the different fabrication methods for the ZnO layers, might also influence on the charge transport of these two types of ZnO devices. However, since this system is complicatedly affected by both the thickness of blend layers and the morphology of ZnO layers, it is difficult to confirm the reason for the different charge transport in ZnO NW-based device compared to the solar cell based on ZnO seed layer.
Despite the unclear charge transport and recombination mechanisms that can be further investigated in the future, it is clear that the ZnO NW can improve the device performance with great potential to further enhance the efficiency.